January 1, 1911.] 



THE INDIA RUBBER WORLD 



121 



Specifications for the Purchase of Materials — I. 



By Frederic Damierth, Ph.D.* 



PROBABLY the most important work of the chemical engi- 

 neer consists in devising methods of testing and finding 

 out the various properties of materials which render them 

 valuable and useful ; and furthermore the assembling of these 

 various properties in the form of specifications for purchasing. In 

 this important phase of his work he is aided to a considerable 

 extent by the ground which has been covered in years past by 

 our large railway corporations and by the deliberations of such 

 representative bodies as the American Society for Testing 

 Materials, the National Fire Protection Association, and the 

 International Congress of Applied Chemistry. Last, but not 

 least, should be mentioned the extensive specifications for deliv- 

 ery which have been issued by the several departments of our 

 own as well as foreign governments. All the literature issued 

 by the specification bureau of a large corporation does not 

 necessarily pertain directly to this subject, although it is a prod- 

 uct of the observations made by that bureau. The five more 

 important types of publications thus issued may be defined as 

 follows: (1) Specifications showing properties required, by the 

 purchaser, of materials to be delivered. 



(2) Methods for making the chemical analyses and deter- 

 minations referred to in the specifications. 



(3) Methods to be followed in the purchase of supplies. 



(4) Instructions to producers and supply houses as to the 

 size of packages, character of the containers, and the mode of 

 delivering supplies. 



(5) Instructions to employes, showing methods of mixing in- 

 gredients for paints, etc. 



(6) Instructions to employes, showing precautions to be ob- 

 served in the use of supplies, to prevent waste. 



In an address delivered several years ago before one of our 

 national institutes the late Dr. Dudley in discussing the plan 

 followed by him and his associates said, "First try to find out 

 what you want." This desire arises from the fact that some 

 difliculty has arisen in the processes or the service of a machine, 

 some parts of constructions fail, or some material at present in 

 use does not give satisfactory results, and an investigation is 

 made to see if the cause of the difficulty can be located ; or some 

 product which is being largely used and which is being furnished 

 by different makers, is believed not to be of equal quality from 

 the different sources ; or it is desired to standardize certain prac- 

 tices and make them uniform in all the mills of the corporation. 

 As an element in this last problem, the same quality of material 

 must be furnished and used. 



These and other elements lead us to make investigations into 

 the nature of the commercial products involved, and finally re- 

 sult in specifications. The specifications, after being made, are 

 placed in the hands of a purchasing agent, and by him are made 

 a part of the contract on which materials are bought. Ship- 

 ments of material in accordance with this contract being received, 

 each shipment or specified definite amount is sampled, and the 

 samples examined in accordance with the specifications. If 

 the samples stand the test, the material is "accepted" and paid 

 for; if not, the material is "rejected" and returned to the 

 makers. 



The question now arises : "How shall a specification be 

 made?" It is very obvious, after a casual observation of the 

 case, that in every specification two parties are primarily inter- 

 ested — the producer of the material and the buyer or consumer. 

 It has in fact been said that a specification was an attempt on 

 the part of the consumer to tell the producer what he wanted. 

 As a result of this view, our early specifications did little more 



•Consulting Chemical Engineer, Philadelphia. 



than define the "qualities of the material." Our modern specifi- 

 cations go beyond this and indicate the method of sampling, 

 state how much material one sample shall represent, and pre- 

 scribe methods of testing either in whole or in part. The early 

 antagonism of the producer and the consumer later gave way to 

 more weighty questions such as: How much material shall be 

 involved in one test? Would it not be advisable to buy in lots 

 of the same size as the test involves? How shall samples be 

 drawn? Shall the quality of the shipment be determined at 

 the mill where the material is made, or shall it be done after the 

 shipment is received? In the former case an inspector (rep- 

 resentative of the purchaser) must be employed. 



In many cases it is found that those who ultimately receive 

 the material, such as the storekeeper, or the foreman of a de- 

 partment, can frequently make certain inspections better and 

 cheaper than any one else. Accordingly they must be supplied 

 with proper instructions. 



It being thus necessary to incorporate information for a num- 

 ber of different parties, the modern specification has in many 

 instances become rather lengthy, and at times apparently un- 

 wieldy. The desire on the part of some manufacturers to de- 

 liver the lowest possible quality on a contract, and the endeavor 

 of the testing chemist and testing engineer to enforce the con- 

 ditions of the specification, frequently give rise to the wish that 

 the specifications had been even more definitely worded. In 

 preparing any specification, the aim should be to incorporate in- 

 formation which may be needed by (1) the manufacturer, (2) 

 the testing chemist, the testing engineer, the inspector, and (3) 

 the persons who are to use the material. On the other hand, 

 care should be taken to eliminate all facts or statements which 

 have no direct reference to the specification in hand. 



The remarks just made will no doubt be followed by such 

 questions as: "Should the specification cover the chemical 

 method involved in the analysis of material?" "should it cover 

 the methods of testing?" "should it include complete instruc- 

 tions for the inspector, covering every point ?" These questions 

 may be answered by the following statements, or rules : 

 .A specification should not attempt too much nor be too compli- 

 cated. Well known methods of analysis or methods of test- 

 ing should be referred to only in a general way. 

 Well known precautions to be observed by the inspector should 



not be explained in detail. 

 New methods of analysis and testing, not well known, must be 

 described in detail, or reference must be made to the original 

 publication. 

 Analyses such as are not conducted in the same manner by all 

 chemists should be described and issued in separate form 

 and made a part of the specifications. 

 When drawing up a specification, do not incorporate in it all 



that you know about that particular subject. 

 Do not put too many restrictions into the specifications; and 

 state as few tests as are necessary to yield the product 

 required. 

 Do not make the limits too severe. If you tie the manufacturer 

 down to the extreme limit, you place yourself in a position 

 of absolute antagonism to him. 

 It is better to specify a good average material, and get the neces- 

 sary protection in machines and processes by a more liberal 

 factor of supervision and technical knowledge, than to in- 

 sist on extreme limits, which can only lead to constant fric- 

 tion, and a demand for concessions. 

 Do not think that the most perfect specimen of a certain product 

 should be made to represent the total output of the works. 



